platform engineer for multi-agent systems

EPAM develops enterprise software products, open source solutions, and accelerators. The role focuses on an enterprise AI agent development platform that enables global-scale AI orchestration within highly secure environments.

• Own the AWS AgentCore Gateway architecture, focusing on A2A protocol routing and enforcement;
• Design deployment topologies for multi-tenant and single-tenant AgentCore Gateway models;
• Implement routing mechanisms and trust models for agent discovery and invocation lifecycle;
• Integrate and enforce OAuth 2.0 and SigV4 authentication for A2A communication channels;
• Develop backend features in Python leveraging the AgentCore SDK for gateway interaction;
• Perform performance benchmarking and security hardening for critical gateway workloads;
• Collaborate with platform, security, and orchestration teams on governance models for multi-agent environments;
• Support enterprise-scale deployment, monitoring, and resilience of A2A traffic flows.

• 5+ Years of experience in platform or backend engineering roles;
• Hands-on production experience with agent-to-agent (A2A) or agent-to-tool gateway platforms;
• Strong knowledge of OAuth 2.0 authentication and identity-driven routing models;
• Proven experience implementing machine-to-machine trust models in production systems;
• Practical understanding of multi-agent system architectures and invocation lifecycle;
• Proficiency in Python backend development for platform-level services;
• Familiarity with AWS-native architectures and deployment strategies for cloud-native systems;
• Nice to have: Direct experience with AWS AgentCore Gateway and its A2A enforcement model, familiarity with AWS AgentCore Runtime multi-agent invocation patterns, background in authoring enterprise A2A governance standards and best practices, hands-on experience with LangGraph or CrewAI for multi-agent orchestration at scale.

• Remote work in Portugal.